The work in this area attempts to formulate a broad conceptual framework within which one can incorporate as much of the phenomonology of carcinogenesis as possible, and to develop mathematical models for specific aspects of carcinogenesis within this framework, that lead to testable predictions or suggestions for original experiments. Efforts are currently directed toward both fundamental and applied theory. The former category presently includes models of the host defense interacting with the primary tumor, and models of metastasis formation. In the latter category a specific model for breast cancer epidemiology has been proposed.